Subnational, high-frequency, near real-time Europe GDP data

Our Europe GDP data is released before official statistics, and provides higher frequency insights.

This is a preview for 2015 across ~1300 EU NUTS3 regions with monthly frequency.

We have data in Europe from 2014 – present.

Press the play button on the interactive 3D map to see how GDP changed. Click, hold, and drag the right mouse button to change viewing angle. Scroll to zoom in.

Customers/ partners/ affiliations:

Our Europe data provides more real-time and higher frequency insights than official data

Make better decisions with more granular and real-time economic insights

Some examples of how our Europe data has been used by customers/ partners across industries.

Our models' predictions are ~98% similar to official statistics

We compared our models’ prediction of annual GDP in 2019 to official data in 2019. Our predictions are ~98% similar to official stats across ~1300 regions. Use the dashboard to see our accuracy rates per country and per region.

Leverage economic knowledge from the world's #1 institution

How it works

We use satellite imagery, machine learning, luminosity, and other inputs to calculate subnational GDP.

Our department is ranked #1 in Economic Geography worldwide, and our methods build off an approach from the American Economic Review, the world’s top Economics journal. We were funded by the European Space Agency BIC UK.

We’re a team of Fellows, PhDs, and postdocs from the LSE whose research specializes in this field, and we’re bringing our findings into industry.

In benchmarking tests in the EU, our methods average ~98% accuracy.

We use NASA satellite imagery, luminosity,  machine learning, and other inputs to estimate GDP.

We specialise in spatial economics data

We are Fellows/ PhDs/ Postdocs from the London School of Economics, initially funded by the European Space Agency BIC UK. We have published in peer-reviewed journals, are Forbes 30 under 30 entrepreneurs, and designed policies for national governments.  We are complemented by ex-SoftBank and ex-McKinsey advisors.

PhD Economic Geography, LSE

Visiting Fellow, LSE

Forbes 30 under 30

PhD Economic Geography, LSE


Download European monthly 2015 subnational data for free

Download data for 25 countries and 1297 regions for each month in 2015. You get 15,564 rows of data, broken down into the following fields.

nuts0Admin level 0 (country-level, highest level of granularity)ItalyString
nuts1Admin level 1NORD-OVESTString
nuts2Admin level 2PiemonteString
nuts3Admin level 3 (lowest level of granularity)CuneoString
GDP_nuts3GDP of Admin level 3 in millions of euros1519.358Float
MoM_GDP_Growth_nuts3Percentage change of Admin level 3 GDP from last month1.27String
YoY_GDP_Growth_nuts3Percentage change of Admin level 3 GDP from last year3.07String
GDP_nuts2GDP of Admin level 2 in millions of euros10768.4Float
MoM_GDP_Growth_nuts2Percentage change of Admin level 2 GDP from last month1.14String
YoY_GDP_Growth_nuts2Percentage change of Admin level 2 GDP from last year2.95String
GDP_nuts1GDP of Admin level 1 in millions of euros46024.89Float
MoM_GDP_Growth_nuts1Percentage change of Admin level 1 GDP from last month1.23String
YoY_GDP_Growth_nuts1Percentage change of Admin level 1 GDP from last year3.21String
GDP_nuts0GDP of Admin level 0 in millions of euros140536.4Float
MoM_GDP_Growth_nuts0Percentage change of Admin level 0 GDP from last month1.02String
YoY_GDP_Growth_nuts0Percentage change of Admin level 3 GDP from last year2.91String
id_nuts0Unique identifier for nuts 0 (e.g. UK)ITString
id_nuts1Unique identifier for nuts 1 (e.g. UKC)ITCString
id_nuts2Unique identifier for nuts 2 (e.g. UKC1)ITC1String
id_nuts3Unique identifier for nuts 3 (e.g. UKC11)ITC16String
timeUnique identifier for time series13Integer

Preview data

Our Europe data provides more real-time and higher frequency insights than official data. 1300 regions, 25 countries, 2014 – present.

Our Global data provides subnational economic understanding when many economies only have national-level information. ~60,000 regions, 195 countries, 2000 – 2020.

Our Luminosity data isolates for human-generated light better than alternatives. Millions of luminosity points globally. 2014 – present.